#280: Medication ontology orgainzed by ingredient
-------------------------+---------------------------
Reporter: preeder | Owner: rwaitman
Type: enhancement | Status: assigned
Priority: major | Milestone: morning-star
Component: data-stds | Resolution:
Keywords: | Blocked By:
Blocking: |
-------------------------+---------------------------
Changes (by dconnolly):
* cc: gpc-dev@… (added)
* owner: dconnolly => rwaitman
* status: new => assigned
* priority: minor => major
Comment:
Russ,
We'd like your take on the [ticket:199#comment:12 UNMC May 15 proposal];
it discussed in our 9 Jun teleconference (#12)
- a
[https://drive.google.com/open?id=0B02nuw10QdWQem5yajdQN0RJOG8&authuser=0
recording of the UNMC meds demo and discussion] was captured
Much of the discussion was relevant to organizing meds by ingredient:
PN: the organization looks nice
PR: I’m mostly concerned about stability and deployment cost
DC: any studies on the horizon that use this stuff?
JRC: perhaps adaptable? I haven’t seen the SQL they’re going to use; i.e.
which CDM elements they’re going to exploit.
JRC: as a clinician, I’d look for ADAPTABLE sorts of stuff in Current med
reconciliation data in outpatient order data
SK/UMN: Q1: where is ATC hierarchy derived from?
JRC says NLM but they provide other options as well.
Q2: Are the NDC folders in i2b2 named after NDC codes?
DC: Does everything in formulary have an NDC code?
JRC: unknown. we haven't done in-patient meds data. (current terms have
about @27k NDC, 5100 RXNorm codes)
DC: What if we do a text search for warfarin? Is it easy to see “the right
thing” in the results?
JRC tries metroprolol); gets “too many results” initially, but then once
we get past that, yes, it’s clear which is “the right one” in the results.
DC: how difficult to implement? HH: metadata build takes roughly 2 hours,
but they are using API and have been IP whitelisted for high use (?)
NG: At KUMC in HERON, we get few hits on NDC codes from our formulary
JRC: low hits perhaps due to historical NDC codes not supported, NLM has
promised to support historical codes in their online web service.
DC to get review by Russ
To PR's point about stability and deployment cost and the possibility of
re-visiting the GPC 1.0 meds ontology (#78), I realize we already
discussed that 21 Apr (#12), resulting in this ticket, #280.
#280 is currently in "the someday pile"; I tried to figure out if any of
our plans depend on this, but I haven't found anything in particular. In
the 15 May proposal, JRC writes:
CDM V3 Dispense and Prescribe Tables require medication data elements
not represented in what is rolled out at KU.
I'm still a little fuzzy on which data elements those are and whether
they're required to answer research questions of interest.
--
Ticket URL:
<http://informatics.gpcnetwork.org/trac/Project/ticket/280#comment:3>
gpc-informatics <http://informatics.gpcnetwork.org/>
Greater Plains Network - Informatics
_______________________________________________
Gpc-dev mailing list
[email protected]
http://listserv.kumc.edu/mailman/listinfo/gpc-dev
